An integer programming model for the selection of pumped-hydro storage projects

TL;DR

This paper presents an integer programming model for selecting pumped-hydro storage projects using digital elevation data, optimizing for cost and feasibility to support renewable energy integration.

Contribution

It introduces a novel optimization model that integrates digital elevation models for efficient PHS siting and proposes solution heuristics to reduce computational complexity.

Findings

Successfully identified cost-effective PHS sites near Sobradinho, Brazil.

Demonstrated the model's ability to incorporate various constraints and cost factors.

Provided a scalable heuristic for reservoir closure and site selection.

Abstract

Energy storage systems - in particular, Pumped Hydropower Storage (PHS) - will be increasingly important to support the transition of power systems toward zero emissions. The reason is that PHS can mitigate the variability and uncertainty of renewable energy production from solar and wind power to balance electricity demand with supply. In this paper, we propose an integer programming problem for PHS siting that uses a Digital Elevation Model (DEM) to meet an energy storage requirement. It assumes the existence of a reservoir, lake, or river, and decides where to build a reservoir that will constitute the PHS with the existing body of water. This model finds minimum-cost project candidates given parameters such as desired head, power, and operation time. The paper discusses different solution methods to assure reservoir closure and avoid its fragmentation. A heuristic explores the representations of the DEM, from more aggregate to more precise, to sequentially refine the solution based on the last selected site, which reduces computational effort. The formulation is general and the objective function includes both construction and equipment costs. Constraints are related to the energy storage target and reservoir closure based on the DEM. We illustrate the methodology by selecting multiple PHS projects next to the reservoir of the Sobradinho hydropower plant in Brazil. The result of this model can be seen as a bottom-up step that prepares PHS candidate projects to be considered by an integrated resource planning model in a top-down step, that would select from these shortlisted projects.